Chat GPT, the cutting-edge artificial intelligence chatbot that generates content, was developed by Open AI in November 2022. It gathered 100 million active users, just two months after its launch and, over the last year, it has taken the professional world by storm. But, do you know the ChatGPT risks that this AI software poses to your business? Keep reading to learn more about what the risks of ChatGPT are and how your business can avoid them.
What is ChatGPT?
It can answer almost any question and generate articulate content on a seemingly infinite range of subject areas, leading to many businesses using the artificial intelligence-powered chatbot to complete administrative tasks, write long-form content like letters and essays, create job specifications, and much more.
Despite its extraordinary success and obvious benefits, the increased use of ChatGPT in the workplace has left many to question its legal and compliance implications for businesses. According to a survey, two-thirds of organisations in the UK are either implementing or considering a ban on ChatGPT and other generative AI applications within the workforce. Several high-profile companies, including Amazon and Goldman Sachs, have already banned the use of ChatGPT in the workplace in an attempt to avoid the risks presented.
What are the Risks of ChatGPT?
We outline some of the issues that your business should be aware of when using ChatGPT and provide some tips to get ahead of the risks associated with it.
1. Inaccuracy
The ChatGPT interface will openly inform you that it may not be accurate. Since it relies on human feedback and information found on the internet, there is an inherent risk that users will receive inaccurate, incomplete or biased information.
There have indeed been many instances where ChatGPT has produced downright wrong responses. Two US lawyers were subject to fines as they used ChatGPT to draft arguments for an aviation injury claim that included fake court citations. This is obviously a risk for your business if an employee shares inaccurate information with a customer or client.
It is, therefore, crucial that employees using ChatGPT for work-related projects do not take answers at face value and are trained to independently check any responses provided, including financial and market information. It is also worth employers completing a spot-check of the data your organisation is using from ChatGPT to ensure it has been independently verified. If the case of the AI lawyer is anything to go by, AI from a legal perspective, has a very long way to go.
2. Confidentiality and Data Protection Risks
Despite only being in existence for less than a year, ChatGPT has already resulted in many users inadvertently disclosing businesses’ confidential information, trade secrets and other propriety information. It is still unclear to what extent ChatGPT and other AI tools retain and rely on confidential information entered by a user to produce something similar. As a result, if an employee inputs sensitive information in order to draft a client letter, for example, this could potentially be a data breach.
Businesses should, therefore, consider prohibiting employees from inputting confidential or proprietary information into chatbots such as ChatGPT.
3. Discrimination and Bias
Employers have been using ChatGPT and other AI tools to make employment decisions, including the recruitment, monitoring and firing of workers.
Although AI tools remove unconscious human bias from decision-making, they may inadvertently discriminate against employees or job applicants based on objective factors that can usually be explained by protected characteristics.
Additionally, it may unintentionally perpetuate certain stereotypes or reinforce existing biases that are prevalent in society. For example, in 2018, Amazon had to ditch a CV screening algorithm it had been using, as the algorithm (that used Amazon’s own data on previous recruitment decisions) had taught itself that male candidates were preferable to female candidates. This led to CVs being filtered out if they contained the word ‘women’s’.
As a result, businesses using ChatGPT for the recruitment of new staff should be careful about the criteria they set, be aware of any possible issues and constantly monitor to safeguard against possible discrimination.
Businesses should, therefore, consider prohibiting employees from inputting confidential or proprietary information into chatbots such as ChatGPT.
3. Discrimination and Bias
Employers have been using ChatGPT and other AI tools to make employment decisions, including the recruitment, monitoring and firing of workers.
Although AI tools remove unconscious human bias from decision-making, they may inadvertently discriminate against employees or job applicants based on objective factors that can usually be explained by protected characteristics.
Additionally, it may unintentionally perpetuate certain stereotypes or reinforce existing biases that are prevalent in society. For example, in 2018, Amazon had to ditch a CV screening algorithm it had been using, as the algorithm (that used Amazon’s own data on previous recruitment decisions) had taught itself that male candidates were preferable to female candidates. This led to CVs being filtered out if they contained the word ‘women’s’.
As a result, businesses using ChatGPT for the recruitment of new staff should be careful about the criteria they set, be aware of any possible issues and constantly monitor to safeguard against possible discrimination.
Next Steps for Businesses
Advances in ChatGPT and similar AI technologies hold exciting opportunities but there are legitimate concerns that must be addressed.
Businesses will have to decide whether they will ban AI tools in their day-to-day business activities or employment decisions or set parameters around its use at work. For example, it may be appropriate to ban the use of ChatGPT for certain roles or types of work products where it is felt the use of the chatbot is not suitable.
If your business chooses to embrace new AI technologies, you need to be aware of the pitfalls and have a clear policy as to how and when they can be used in your organisation. In such circumstances, your business should review and update existing technology use policies as well as codes of conduct and confidentiality agreements to cover emerging AI tools such as ChatGPT. Such policies should be a written source of guidance on how issues should be handled and should include a description of rights and responsibilities for managers, and employees.
Similarly, employees should be encouraged to be transparent about their use of ChatGPT so that the appropriate verifications can be made. According to a survey, nearly 70% of employees use ChatGPT at work without the knowledge of their managers.
Finally, employers who allow the use of ChatGPT in the workplace should train employees on the proper use of the chatbot as well as the realistic risks associated with it.For more advice on how to reduce risks associated with AI or about amending workplace policies to reflect the use of AI in the workplace, contact our expert Employment law team via email.
Image by frimufilms on Freepik
